Abstract
This paper presents a research and development project undertaken by DNV to develop digital twins of hull compartments on floating assets, wherein inspection findings are collected and documented for analysis and reuse. This process aims to drive increased efficiency in future inspections. Data gathering is performed by intelligent autonomous drones capable of self-navigation in structurally complex compartments. These are coupled with AI-driven algorithms that detect and identify findings such as cracks, corrosion, and deformations. The project serves a dual purpose: firstly, to reduce the need for manned entry into hull compartments, and secondly, to enhance the efficiency of planning and execution of future inspections. This is achieved by ensuring known findings are identified, mapped in the digital twin, and can be easily located and reassessed.
The drones are equipped to perform both close and general visual inspections, although further development in lighting is necessary to support general visual inspections of surfaces in large hull compartments. The surveyor then reviews the inspection findings and identifies actual defects, which contributes to the continuous learning of the detection algorithm. Successful trials have been conducted with both free-flying and tethered drones.
Future development is planned to focus on improving automated inspection planning and visualization for monitoring compartment conditions. Additionally, enhancements are aimed at increasing the capability for anomaly detection, improving drone navigation in structurally complex spaces, and enhancing the capacity for data gathering while in motion.
